电力系统短期负荷预测的高木-关野模型研究

STUDY ON THE TAKAJI-SUGENO MODEL OF SHORT-TERM LOAD FORECASTING FOR POWER SYSTEM

电力系统短期负荷预测在电力系统的运行设计中有重要的意义 ,利用模糊神经网络的方法进行电力负荷预测是国际上近年来很热门的一个方向。本文在传统的 BP神经网络基础上 ,提出了一种短期负荷预测的模糊神经网络模型—高木 -关野模型 ,以某供电局 2 0 0 0年的负荷实测值建立模型 ,进行了负荷预测 ,与实际值进行比较分析表明 ,这一模型应用于短期负荷预测能获得较高的预测精度 。

Short term load forecasting is important in power system operation and design. Neural network is a popular way in the world for load forecasting. On the basis of traditional BP neural network, the model of Takaji Sugeno for short term load forecasting of power system was proposed in this paper. Based on load records of some area, the load forecasting model was made and used to forecast the load value, which was compared with the practical value. Practical examples indicate that the application of FNN to short term load forecasting is feasible and effective, and can produce more accurate results than conventional methods.